Lost in Transcription: Subtitle Errors in Automatic Speech Recognition Reduce Speaker and Content Evaluations

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kadoma, Kowe, Shrivastava, Priyal, Naaman, Mor
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912969306144768
author Kadoma, Kowe
Shrivastava, Priyal
Naaman, Mor
author_facet Kadoma, Kowe
Shrivastava, Priyal
Naaman, Mor
contents Researchers have demonstrated that Automatic Speech Recognition (ASR) systems perform differently across demographic groups. In this work, we examined how subtitle errors affect evaluations of speakers and their content using a preregistered online experiment (N=207, U.S.-based crowdworkers). Participants watched speakers with various accents deliver a talk in which the subtitles were accurate or error-prone. Our results indicate that error-prone subtitles consistently reduce both speaker and content evaluations for all speakers. We did not see disparate impact between the accent groups, controlling for subtitle quality. Taken together, though, the findings of this short paper imply that speakers with accents for which ASR systems perform poorly are likely to be further penalized by viewers with lower evaluations.
format Preprint
id arxiv_https___arxiv_org_abs_2603_15807
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Lost in Transcription: Subtitle Errors in Automatic Speech Recognition Reduce Speaker and Content Evaluations
Kadoma, Kowe
Shrivastava, Priyal
Naaman, Mor
Human-Computer Interaction
Researchers have demonstrated that Automatic Speech Recognition (ASR) systems perform differently across demographic groups. In this work, we examined how subtitle errors affect evaluations of speakers and their content using a preregistered online experiment (N=207, U.S.-based crowdworkers). Participants watched speakers with various accents deliver a talk in which the subtitles were accurate or error-prone. Our results indicate that error-prone subtitles consistently reduce both speaker and content evaluations for all speakers. We did not see disparate impact between the accent groups, controlling for subtitle quality. Taken together, though, the findings of this short paper imply that speakers with accents for which ASR systems perform poorly are likely to be further penalized by viewers with lower evaluations.
title Lost in Transcription: Subtitle Errors in Automatic Speech Recognition Reduce Speaker and Content Evaluations
topic Human-Computer Interaction
url https://arxiv.org/abs/2603.15807